time serialization ready for production

data_preparation/time_series.py

@@ -0,0 +1,33 @@
+import pandas as pd
+import numpy as np
+
+
+def start_end_times(filename):
+  df = pd.read_csv(filename)
+  columnname = "Date"
+  dt = pd.to_datetime(df[columnname], format="%Y-%m-%d")
+  print()
+  print(filename)
+  print("min")
+  print(dt.min())
+  print("max")
+  print(dt.max())
+  return dt.min()
+
+
+def timeframes():
+  start_end_times("data/rpe.csv")
+  start_end_times("data/games.csv")
+  start_end_times("data/wellness.csv")
+
+
+def normalize_time_series(filename, start):
+  df = pd.read_csv(filename)
+  columnname = "Date"
+  dt = pd.to_datetime(df[columnname], format="%Y-%m-%d")
+  df["TimeSinceAugFirst"] = (dt - start).dt.days
+  df.to_csv("cleaned/time_series_" + filename)
+
+
+start = start_end_times("data/rpe.csv")
+normalize_time_series("cleaned/rpe.csv", start)